Deployment optimsation

The Jtrix service model enables access to services to be as uniform as possible across the Jtrix. The service binding mechanism allows services to adapt their implementations to the environment to which they are being deployed, in order to allow this uniformity.

This is useful behaviour at the macro level, where large subsystems are being operated by different parties, but consider a single application, which may itself be composed of many subsystems, each of which is an independent Jtrix entity and may be administered as a separate, bought-in service, or as part of a larger application. It is likely that many of these subsystems will have similar deployment and security constraints.

Ideally, we wish to be able to control and optimize the deployment of netlets when considering aggregations of services and applications, in order to maximise performance by identifying and minimising bottlenecks. Thus must happen at an application level, and it would be nice for it to happen at a higher level, with optimisation being performed across administrative boundaries.

We'd like a mechanism by which we can locate hosting resources which will allow us to optimise our deployment with respect to the other resources we are using.